MACHINE LEARNING MODELS FOR PREDICTING MECHANICAL DAMAGE, VIGOR AND VIABILITY OF SOYBEAN SEEDS DURING STORAGE

Aug 4, 2026

Artificial Intelligence has been widely applied in data prediction for better decision making and process optimization. In the post-harvest, controlling biotic and abiotic factors is fundamental for conserving seed quality. The tetrazolium test helps evaluate quality but has limitations that can cause evaluation errors. Machine learning offers an alternative to predict soybean seed quality, providing faster results than laboratory methods and improving robustness with low operational cost. The study identified the best model to predict mechanical damage, vigor, and viability under different storage conditions and times.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Introduction

Seed storage variables and quality preservation

Mechanical damage and tetrazolium test limitations

Need for faster, data-driven quality assessment

Engenharia Agrícola ISSN: 1809-4430 (on-line) [www.engenhariaagricola.org.br](www.engenhariaagricola.org.br) | Scientific Paper
Doi: [http://dx.doi.org/10.1590/1809-4430-Eng.Agric.v43n4e20230121/2023](http://dx.doi.org/10.1590/1809-4430-Eng.Agric.v43n4e20230121/2023)
MACHINE LEARNING MODELS FOR PREDICTING MECHANICAL DAMAGE, VIGORAND VIABILITY OF SOYBEAN SEEDS DURING STORAGE
Laila R. Cirqueira1, Paulo C. Coradi1,2,3*, Larissa P. R. Teodoro1, Paulo E. Teodoro1, Dágila M. Rodrigues2,3
3*Corresponding author. Laboratory of Postharvest (LAPOS), Federal University of Santa Maria, Campus Cachoeira do Sul, Passo D’Areia/Cachoeira do Sul-RS, Brazil.
[E-mail: paulo.coradi@ufsm.br | ORCID ID:](E-mail: paulo.coradi@ufsm.br | ORCID ID:) [https://orcid.org/ 0000-0001-9150-2169](https://orcid.org/ 0000-0001-9150-2169) | | | | --- | --- | --- | | KEYWORDS
artificial intelligence, postharvest, storage packaging, storage temperature; storage time, tetrazolium test. | ABSTRACT
Artificial Intelligence has been widely applied in data prediction for better decision making and process optimization. In the post-harvest, the control of biotic and abiotic factors is fundamental for the conservation of seed quality. Meanwhile, the tetrazolium test has been used to evaluate seed quality, however, with several limitations that can lead to evaluation errors. Thus, machine learning models can be an alternative to predict the quality of soybean seeds, with gains in the speed of obtaining results in relation to laboratory analysis methods, making the processes more robust and with low operational cost. With this, the aim of this study was to identify the best machine learning model for predicting mechanical damage, vigor and viability of soybean seeds during storage, depending on different conditions (10, 15 and 25 ºC), packaging (with coating and uncoated) and storage times (0, 3, 6, 9 and 12 months) . M5P decision tree (M5P) and Random Forest (RF) models showed the best performance for predicting seed vigor (r = 0.75 and MAE = 10.0), and viability (r = 0.85 and MAE = 5.1), and mechanical damage to seeds (r = 0.64 and MAE = 11.2). It was concluded that the Random Forest (RF) model was the one that best predicted the results of soybean seed quality, with a more simplified and agile analysis for the development of vigor and viability of soybean seeds in storage. | | | INTRODUCTION
The moisture content of the seeds, the temperature and relative humidity of the intergranular air and the storage environment are important variables to be monitored to preserve the quality of the seeds (Capilheira et al., 2019) . However, variations in seed moisture content, shape, environment and storage time can influence the metabolic activity and physiological seeds quality (Mylona et al., 2012) .
To reduce the metabolic activity of the seeds, it is suggested to control the temperature and relative humidity of the storage environment, so that the seeds remain in equilibrium moisture content with moisture content close to 12%(w.b.), considered a safe moisture (Ebone et al., 2020; Sarath et al., 2016). According to Oliveira et al. | | (2021), physical damage caused to seeds can cause reduced vigor, viability and even seed death (Rocha et al., 2017; Silva et al. , 2022) . For this, the tetrazolium test has been an important and efficient analysis to assess quality physical and mechanical damage, which interfere with seed vigor and viability (Rocha et al. , 2017) However, the tetrazolium test has some limitations, including the need for advanced training and knowledge about seed science and technology to interpret the results, with the possibility of susceptible errors (Coradi et al., 2020) . Seed quality analyzes often generate a quantity of information that makes a quick and effective short-term analysis impossible. Therefore, erroneous results may imply economic losses for seed processing units (André et al., 2022) . | 1 Federal University of M

FAQs

Why is controlling post-harvest storage conditions important for seed quality?

Moisture content, temperature, and relative humidity influence seed metabolic activity and physiological quality, so maintaining appropriate conditions helps preserve seed quality.

What limitations does the tetrazolium test have when evaluating seed quality?

The test requires advanced training and knowledge to interpret results, and this can lead to susceptible evaluation errors.

How do machine learning models improve prediction of soybean seed quality during storage?

Machine learning models can predict mechanical damage, vigor, and viability more quickly than laboratory analyses, enabling more robust decision-making with low operational cost.

Aug 4, 2026

Upload by Ava Thompson